基于神经网络的敏捷制造业自动化物流信息预测的研究与实现

Research & Realization on Neural Network for AM Automatic Logistics Information-System's Forecast

提出基于神经网络的敏捷制造系统物流预测模型及其算法,并采用动态反馈算法(DBF)对网络进行了训练。结合实际敏捷制造企业,通过与传统的预测方法进行比较,证明该预测系统具有良好的非线性函数逼近能力和隐层反馈的动态数据处理能力,能够学习与适应严重不确定性系统的动态特性,预测精度高。

This paper presented research on neural network for AM automatic logistics information system＇s modeling and adopted DBP（dynamic back propagation algorithm） to train the network. Contrasted to traditional forecasting methods, the foraecasting system herein has good ability for approaching to non-linear function and latent feedback＇ s dynamic data processing, it can learn and adapt to severity uncertainly dynamic system＇s tract, and the forecasting precision is high.